**** BEGIN LOGGING AT Thu Sep 22 02:59:57 2011 Sep 22 05:13:28 morning Sep 22 07:08:19 RP_, it seems you missed some of the patches (gst-plugins-base, glib-2.0, e2fsprogs, gcc reg. plugin includes, eglibc nativesdk packaging). Was it on purpose? Sep 22 07:15:24 good morning Sep 22 07:27:52 hello all Sep 22 07:30:02 Hello. I'm have problem with compile linux_2.6.30.bb at at91sam9263ek machine. In linux_2.6.30.bb file do not defined folder /at91sam9263ek and do not found file "defconfig".output error: "No such file or directory: '/home/leonopulos/stuff/openembedded/recipes/linux/linux-2.6.30/./defconfig'". How to fix this error? Please, help. Sep 22 07:33:12 instead of ".../stuff/openembedded/recipes/linux/linux-2.6.30/./defconfig" must be ".../stuff/openembedded/recipes/linux/linux-2.6.30/at91sam9263ek/defconfig" Sep 22 07:36:42 OE has been updated yesterday Sep 22 07:41:20 leonopulos, create a linux-2.6.30/at91sam9263ek/ folder and put there your defconfig (probably you can get one in one of earlier kernels). Sep 22 07:43:01 lumag: sorry, but this folder available on disk. Sep 22 07:43:28 and there is a defconfig? Sep 22 07:44:10 and there is defconfig Sep 22 07:44:12 yes Sep 22 07:44:13 looks like the problem is inside linux_2.6.30.bb Sep 22 07:44:20 yes Sep 22 07:44:38 it is there Sep 22 07:45:08 leonopulos: you can switch to an older (and more stable) kernel recipe, or patch the latest Sep 22 07:45:19 no reference to this directory Sep 22 07:47:12 leonopulos: I used to use linux_2.6.28.bb, take a look there Sep 22 07:47:24 mckoan: no, sorry, i need this version. in the previous not AC97 codec Sep 22 07:48:50 mckoan: yes. i myself changed on 2.6.30 Sep 22 07:50:37 leonopulos: we can't figure out how is your new linux_2.6.30.bb, if is changed pastebin it then Sep 22 07:50:43 ~pastebin Sep 22 07:50:43 [~pastebin] A "pastebin" is a web-based service where you should paste anything over 3 lines so you don't flood the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://bin.cakephp.org/ , http://asterisk.pastey.net/ , or install pastebinit with yum or aptitude. Sep 22 07:52:21 ~vim Sep 22 07:52:21 rumour has it, vim is better than emacs Sep 22 07:52:28 xDD Sep 22 07:53:37 mckoan: sorry, thanks Sep 22 07:55:20 morning folks Sep 22 07:57:29 mickey|office: hi Sep 22 07:58:19 mckoan: http://pastebin.com/CGaxkRwq Sep 22 08:03:07 JaMa: ping Sep 22 08:03:23 pong Sep 22 08:03:36 gm, may I PM you to avoid OT? Sep 22 08:04:25 ant_work: I do not care OT traffic if the channel is not busy. Sep 22 08:04:27 y Sep 22 08:05:24 ;) Sep 22 08:06:09 Friday's mood? Sep 22 08:23:42 I get an error "ERROR: Nothing PROVIDES '/home/user/svn/system/openembedded/recipes/grub/grub_0.97.bb". What could be the cause? Sep 22 08:28:00 leonopulos: I said to pastebin your new linux_2.6.30.bb Sep 22 08:31:50 I get an error "ERROR: Nothing PROVIDES '/home/user/svn/system/openembedded/recipes/grub/grub_0.97.bb". What could be the cause? Sep 22 08:38:16 hi all Sep 22 08:45:49 mckoan: yes. now do. Sep 22 08:50:23 i am using a custom built omap3 board and i have downloaded the filesystem for dm3730 evm board and while booting the booting process hangs after priniting the date Sep 22 08:50:26 what could be wrong Sep 22 08:50:35 can anyone help me in debugging this Sep 22 08:52:53 hi all Sep 22 08:53:29 hi bluelightning Sep 22 08:53:46 I was waiting 4 u for quite sometime :) Sep 22 08:54:02 hi Noor Sep 22 08:54:09 oh, what can I do for you? Sep 22 08:54:41 bluelightning: remember yesterday u told me to use _append in BB_HASHBASE_WHITELIST? Sep 22 08:54:56 I think I have found a bug in its implementation Sep 22 08:55:08 I get an error "ERROR: Nothing PROVIDES '/home/user/svn/system/openembedded/recipes/grub/grub_0.97.bb". What could be the cause? Sep 22 08:55:28 not in _append implementation but BB_HASHBASE_WHITELIST Sep 22 08:56:57 bluelightning: the scenario is this when I do BB_HASHBASE_WHITELIST_append in my distro.conf file it appears in my environment (i checked with bitbake -e option ) but the stuff that I have appended does not show up in sstate-cache file Sep 22 08:58:43 further digging the problem I came to know that BB_HASHBASE_WHITELIST value is being picked in SignatureGeneratorBasic calss in bitbake which is present in siggen.py and assigned to self.basewhitelist but what I think is append is not applied at that time Sep 22 08:58:55 Noor: hmm... Sep 22 08:59:04 so self.basewhitelist does not take the updated value Sep 22 08:59:23 I'm not entirely sure it's intended that you append to BB_HASHBASE_WHITELIST anywhere other than where it's populated, it's more of an internal variable Sep 22 08:59:39 for sstate you may have better luck using vardepsexclude Sep 22 08:59:57 How is the update policy these days. Can one expect important updates/patches to recipes to be committed both in oe-classic and oe-core? Sep 22 09:00:55 kobe_: could you pastebin the whole message? Sep 22 09:02:58 mckoan: http://pastebin.com/0QUi0c3r Sep 22 09:04:38 morning Sep 22 09:04:40 bluelightning: http://pastebin.com/KFC4QEED Sep 22 09:04:59 bluelightning: will play with vardepsexclude but I think BB_HASHBASE_WHITELIST needs an update ..... will discuss this with kergoth and see if we comeup with the plan .... coz i saw BB_HASHBASE_WHITELIST at http://www.digipedia.pl/usenet/thread/18578/363/ and it does not say that one should not use it Sep 22 09:06:46 kobe_: FYI, newer versions of bitbake will report exactly why this kind of situation occurs Sep 22 09:07:29 kobe_: what's your MACHINE set to? Sep 22 09:07:31 hi all Sep 22 09:07:42 x86 Sep 22 09:08:01 I'm searching for imgloader and mkimg (following the AMD Embedded) ... who is cosmicpenguin ? :) Sep 22 09:09:06 kobe_: "x86"? is that something you've defined yourself? AFAIK that's not normally a valid value for MACHINE Sep 22 09:09:54 tasslehoff: expect oe-classic read only soon Sep 22 09:10:03 yes, that's defined myself Sep 22 09:10:18 other recipes don't give problems Sep 22 09:11:10 kobe_, how you install your kernel+rootfs on your x86 board ? Sep 22 09:11:52 currently I"m still using syslinux, but I want to move to grub Sep 22 09:12:06 kobe_, any glue on how to ? Sep 22 09:12:20 once I have a kernel and a rootfs, what's next ? Sep 22 09:12:33 hi Sep 22 09:12:41 ant_work: soon, or really soon? Sep 22 09:14:21 alx: for syslinux, I made two partitions: 1 FAT32 and 1 EXT3. On the first partition I install syslinux and the kernel; on the secon partition, I install the rootfs Sep 22 09:14:30 afaik back in June it was decided there would'nt be any new release and the repo would have been closed soon Sep 22 09:14:48 kobe_: for fun, try kexecboot Sep 22 09:14:52 kobe_, by hand, or with a tool ? Sep 22 09:15:13 alx: currently still by hand Sep 22 09:15:33 :( ok Sep 22 09:15:58 kobe_, going same way :( Sep 22 09:16:04 kobe_: ok... well the grub recipe has an explicit check on HOST_SYS and I guess your machine is not matching it Sep 22 09:16:20 bluelightning, ? Sep 22 09:16:50 alx_: that's why he's getting the "nothing provides" error above Sep 22 09:17:00 bluelightning, for grub ? Sep 22 09:17:05 alx_: yes Sep 22 09:17:19 bitbake grub-native or bitbake grub ? Sep 22 09:17:30 same for syslinux ... native or not ? Sep 22 09:18:54 ant_work: ok. will have to speed up the transition then :) Sep 22 09:19:25 tasslehoff, so syslinux is going to die inside OE ? Sep 22 09:20:31 bluelightning: Is it the COMPATIBLE_HOST variable? Sep 22 09:21:15 kobe_: no, it's the code in the "python __anonymous" function within the grub recipe Sep 22 09:21:33 alx_: um, was that for me? Sep 22 09:22:41 Mm, I see no python functino in the recipe Sep 22 09:29:07 tasslehoff, no sorry. Sep 22 09:29:38 tasslehoff, I was following oe-classic chat with ant_work. I'll go mad converting oe-classic to oe-core :( Sep 22 09:47:37 what is the idea behind -nativesdk? When I build zlib-nativesdk in an environment for an ARM machine, this sdk is built for x86_64 (e.g. libz.so is for x86_64, not for arm) Sep 22 09:48:01 how can I create sdk's with content for the target platform? Sep 22 09:50:11 kobe_: right, sorry this is OE classic; so in the OE classic recipe it's COMPATIBLE_HOST as you suggested (sorry) Sep 22 09:50:38 alx_: it's not too difficult Sep 22 10:00:45 has anybody worked on OE for nc600 pollux .... Sep 22 10:02:33 03Tristan Lelong  07master * r9e1aa4d02a 10openembedded.git/recipes/libroxml/ (libroxml.inc libroxml_2.1.1.bb): (log message trimmed) Sep 22 10:02:33 libroxml: Add version 2.1.1 (initial recipe) Sep 22 10:02:33 libroxml is a small, fast and powerful xml library. Sep 22 10:02:33 • built successfully using org.openembedded.dev with DISTRO=minimal and MACHINE = "i686-generic" Sep 22 10:02:33 • built successfully using org.openembedded.dev with DISTRO=minimal and MACHINE = "ppce500v2" Sep 22 10:02:33 • tested using projects unit tests on powerpc binaries Sep 22 10:02:34 Signed-off-by: Tristan Lelong Sep 22 10:16:18 how can i update the opkg database files Sep 22 10:26:41 thanks1 Sep 22 10:34:48 rsv: if you mean the index, bitbake package-index should do it Sep 22 10:36:23 bluelightning: i meant the filesystem on the target Sep 22 10:36:45 rsv: er, opkg update ? Sep 22 10:37:17 it updates. but when i try to install the package it cant find it Sep 22 10:42:30 rsv: generally those sorts of issues result from architecture mismatches Sep 22 11:41:56 after todays git pull, external-csl-toolchain is needed to build image? Sep 22 11:42:17 has someone else seen this ^ Sep 22 11:42:20 and also: ERROR: external-poky-toolchain: Recipe file does not have license file information (LIC_FILES_CHKSUM) Sep 22 11:42:23 ERROR: Licensing Error: LIC_FILES_CHKSUM does not match, please fix Sep 22 11:43:19 probably caused by: ERROR: Multiple .bb files are due to be built which each provide virtual/libiconv (/OE/shr-core/openembedded-core/meta/recipes-core/eglibc/eglibc_2.13.bb /OE/shr-core/openembedded-core/meta/recipes-core/meta/external-csl-toolchain_2008q3-72.bb). Sep 22 11:43:30 and the same for virtual/libc and linux-libc-headers Sep 22 11:43:31 Is there a best-practice for creating layers, or should I just follow style? create a conf/layer.conf and organize like I want inside "meta-tasslehoff"? Sep 22 11:44:36 I think that the external stuff could live in a separate layer Sep 22 11:44:45 JaMa: ^^ Sep 22 11:44:58 tasslehoff: yes, try to use .bbappend where you can Sep 22 11:45:14 to minimize deltas Sep 22 11:47:24 ant_work: yep. first step is to make it work for my own recipes, then I'll try to add my ugly patches in a nice way. Sep 22 11:48:04 tasslehoff: look at http://cgit.openembedded.org/cgit.cgi/meta-handheld/tree/ which puts together all sort of trickery (thx bluelightning) Sep 22 11:52:12 ant_work: that's the one I looked at when making mine :) Sep 22 11:52:18 hi eFfeM Sep 22 11:52:31 hey ant_work, all Sep 22 11:52:35 how are you doing? Sep 22 11:53:23 preparing to travel, lot of visits in autumn Sep 22 11:53:59 that remidns me, still have to book for elc Sep 22 11:56:18 I'll try to be in Prague, still unsure Sep 22 12:26:23 What have I (not) done? : WARNING: No bb files matched BBFILE_PATTERN Sep 22 12:26:26 I've just registered & booked my flight Sep 22 12:26:36 looking forward to it :) Sep 22 12:27:30 my boss still needs to approve, probably will stalk him about it tomorrow or so Sep 22 12:27:39 there is no program announced yet, is it? Sep 22 12:27:43 tasslehoff: you'll get that if for example the BBFILE_PATTERN for your layer includes .bbappend files and you don't have any Sep 22 12:28:47 eFfeM: don't think it has been yet no Sep 22 12:30:55 bluelightning: a programme makes it easier to convince mgmt that they should pay for it Sep 22 12:31:38 eFfeM: hmm... I can't imagine it will be too long before it's announced Sep 22 12:32:08 obviously you can give your management some idea of what might be covered by past year's schedules Sep 22 12:33:03 for many conferences though much greater value can be derived from the "hallway track" than actual presentations Sep 22 12:37:40 true and it is already being considered but someone still needs to say OK Sep 22 12:37:58 undoubtly it'll be arranged Sep 22 12:42:43 bluelightning: Found another problem with combo-update Sep 22 12:43:01 bluelightning: http://paste.debian.net/131567/ Sep 22 12:46:57 otavio: so what is the problem exactly? Sep 22 12:47:14 I noticed the 2011-1 branch but seems it is not fully merged into master. Sep 22 12:47:24 bluelightning: same as before ... Sep 22 12:48:33 [12:47:17] Applying: openembedded-core: distro tracking: Update Distro Aliases Sep 22 12:48:34 error: patch failed: openembedded-core/meta/conf/distro/include/distro_tracking_fields.inc:3438 Sep 22 12:48:36 error: openembedded-core/meta/conf/distro/include/distro_tracking_fields.inc: patch does not apply Sep 22 12:48:39 Patch failed at 0001 openembedded-core: distro tracking: Update Distro Aliases Sep 22 12:49:05 bluelightning: that I had, but it still don't work. perhaps there's some change I need to do in the recipes themselves. I've just copied my old ones from oe-classic Sep 22 12:49:23 tasslehoff: not that would cause that Sep 22 12:49:36 tasslehoff: if this is an issue, it's of configuration Sep 22 12:50:24 otavio: well, if I git-am that patch by hand it works fine... Sep 22 12:50:29 or, maybe I have mismatched "'s in my BBFILES-variable :) Sep 22 12:51:00 tasslehoff: ah... :) Sep 22 12:51:31 otavio: (after editing the "openembedded-core" out of the path of course) Sep 22 12:53:08 (devel)~/hacking/ossystems/embedded-linux% git am /home/otavio/hacking/ossystems/embedded-linux/patch-ed9e25e7-00be-416e-9dfd-4ce7e986ec0d/openembedded-core/0002-distro-tracking-Update-Distro-Aliases.patch Sep 22 12:53:12 Applying: openembedded-core: distro tracking: Update Distro Aliases Sep 22 12:53:14 error: patch failed: openembedded-core/meta/conf/distro/include/distro_tracking_fields.inc:3438 Sep 22 12:53:49 bluelightning: now the errormessages went away, but I still get NoProvider for my recipe. this feels like when I started to use oe-classic a couple of years ago :) Sep 22 12:54:52 tasslehoff: assuming the noprovider is not coming up for some other reason (and recent versions of bitbake will report if that is the case) there is still something wrong with your layer config Sep 22 12:55:02 ah, but suddenly it works... Sep 22 12:55:09 suddenly? Sep 22 12:56:14 bluelightning: can't think of one file I've saved since I got that message, but now it works. I must have done something.. Sep 22 13:17:38 bluelightning: my fault; found it Sep 22 13:35:47 otavio: what was the problem? Sep 22 13:52:08 bluelightning: I had 4.7.4 version of qt commited and then it conflicted ;-) Sep 22 13:52:11 bluelightning: stupid me Sep 22 13:54:00 otavio: ah, I see.. that makes sense :) Sep 22 13:54:10 (why it failed I mean) Sep 22 13:55:05 I wonder if we ought to check if the current branch is ahead of the remote tracking branch Sep 22 14:15:58 bbl! Sep 22 14:32:41 bluelightning: the best would be to support the 3way merging Sep 22 14:32:50 bluelightning: but this is not trivial to do so Sep 22 14:33:08 otavio: well, git am has an option to enable that Sep 22 14:33:23 bluelightning: but it requires the base hash to match against Sep 22 14:33:45 ah yes, you're right... Sep 22 14:33:53 leonopulos: please note that in conf/machine/at91sam9263ek.conf there is a default setting of PREFERRED_VERSION_linux = "2.6.28" Sep 22 14:34:11 leonopulos: if you switch to a more recent kernel is at your risk Sep 22 14:36:13 mckoan: Russak saying "nothing ventured, nothing gained" Sep 22 15:10:00 | sed: can't read /home/otavio/hacking/ossystems/embedded-linux/tmp-eglibc-eglibc/work/i586-oe-linux/external-csl-toolchain-2008q3-72-r1/external-csl-toolchain-2008q3-72/elf/ldd.bash.in: No such file or directory Sep 22 15:10:04 | ERROR: Function 'do_configure' failed (see /home/otavio/hacking/ossystems/embedded-linux/tmp-eglibc-eglibc/work/i586-oe-linux/external-csl-toolchain- Sep 22 15:10:20 someone having problem compiling it in today's tip? Sep 22 15:11:26 otavio: are you building external-csl-toolchain? Sep 22 15:11:40 denix0: I didn't intend too Sep 22 15:12:15 so why is it in your dependency tree? can you run bb with -g and check what pulls it in? Sep 22 15:12:28 * otavio checks Sep 22 16:01:13 Crofton, mickeyl, florian: I'm around, just ping me - the regular place in 1 hour? Sep 22 16:04:32 denix: ok great... yes about one hour left Sep 22 16:40:35 whats the best way to get bitbake to spit out a message to the build log for debugging purposes? Sep 22 16:40:52 ive tried bb.debug(#,"msg") without luck Sep 22 16:41:32 futhermore, where do bb.debug messages go? Sep 22 16:41:34 some log? Sep 22 16:45:12 msm: bitbake -D Sep 22 16:45:24 yea, that's just broken then i suspect Sep 22 16:45:42 -D 's add nothing more Sep 22 16:47:19 its driving me nuts =) Sep 22 16:49:37 hmm, it appears not to be working at all :/ Sep 22 16:58:57 denix0, yeah Sep 22 17:05:50 hi all Sep 22 17:07:23 is there any wiki/doc for starting with oe-core ? Sep 22 17:09:41 alx_1: Did you search for oe-core in the Wiki? Sep 22 17:10:17 Sep 22 17:10:33 used to google "oe-core" Sep 22 17:10:34 :) Sep 22 17:10:51 Otherwise the documentation is pretty sparse and you should start by using for example the Ȧngström setup scripts. Sep 22 17:11:20 Sep 22 17:11:50 PaulePanter, the last one seems oe-classic ... Sep 22 17:11:52 isnt' it ? Sep 22 17:12:11 I thought they were switched to oe-core. Sep 22 17:12:57 The Yocto site should also have some information. I hope the others can tell you more. In the repository should also be more information. Sep 22 17:13:18 PaulePanter, that's difficult to learn Sep 22 17:24:06 PaulePanter, http://www.yoctoproject.org/docs/1.1/yocto-project-qs/yocto-project-qs.html ? Sep 22 17:35:53 hi, I'm using arm-angstrom-linux-gnueabi-g++ (GCC) 4.3.3 does any one knows at which version cstdint (-std=c++ox) because it failed Sep 22 17:35:54 error: cstdint: No such file or directory Sep 22 17:36:19 I will update to have a newer version of the meta tool Sep 22 17:37:18 alx_1: Did that document help? You should maybe add "git clone" to your search as this should be written in a how-to/tutorial. Sep 22 17:37:33 alx_1: I have to admit I am still using OE-classic. Sep 22 18:26:43 PaulePanter, btw on oe-classic, if I want to extract a SDK ... how to ? Sep 22 19:24:19 so many dead links on the open embedded wiki Sep 22 21:36:31 hi all Sep 22 21:37:05 how to extract an sdk ? Sep 22 22:25:07 JaMa: last 3 oe-core patches about xorg need to be reverted to apply your patchset clean Sep 22 22:25:11 (http://patches.openembedded.org/bundle/otavio/xf86-fixes/) Sep 22 22:25:15 just fyi Sep 23 00:00:12 so I am using oe-classic, was using version of 2008.1. I am pulling it forward to 2010.x and just did a successful full system rebuild and the size has gone up by an order of magnitude Sep 23 00:00:30 I went from 75-ish MB to 750MB... not sure where that came from Sep 23 00:05:44 There are two bins where it came from: ~160MB is in libc.a in /usr/lib - why is this that big? Then there are a ton of .debug folders - how do I not include these? Sep 23 00:06:55 edit - some text was clipped in the last line: 160MB is in libc.a Sep 23 00:07:04 the .debug folders add up to several hundred MB Sep 23 01:42:00 Anyone about? We're getting some funny network issues with Gumstix - looks like the Gumstix is too quiet so the routers don't even know it's there and connections fail Sep 23 01:42:15 as soon as then Gumstix has initiated connections, in-bound connections work Sep 23 01:42:34 but even then we have occasional connect-failures Sep 23 01:42:42 (which look different) Sep 23 01:50:56 any possibility of duplicate MAC addresses? that can introduce random problems Sep 23 02:22:17 notespace: I don't think so. We did have some duplicate IP addresses on the net last week, but that caused all kinds of mayhem Sep 23 02:22:25 it's gfixed now, of course Sep 23 02:24:06 cjheath: ok, good to hear. Sep 23 02:24:54 I was wondering whether I should configure a 60-second ping to a random address, to keep the ARP tables populated **** ENDING LOGGING AT Fri Sep 23 02:59:56 2011